Phase separation in Lipid Bilayer Vesicles under tension

Mohammad Jibin

Indian Institute of Technology Madras, Adyar, Chennai, Tamil Nadu, 600036

V.A. Raghunathan

Raman Research Institute, C.V Raman Road, Bengaluru, 560080

1. Abstract

Lipids are biomolecules with a hydrophilic head group and hydrophobic tails. They are soluble in non-polar solvents and insoluble in polar solvents. When they are exposed to a polar medium, they arrange themselves into a structure which reduces hydrocarbon chain interaction with solvent so as to minimize the free energy. The presence of two hydrocarbon chains forces lipids to form bilayers. Some multi-component bilayers are known to exhibit phase separation below a critical temperature. Phase separation has recently been observed in a two-component vesicle under tension. This tension can be simulated by stretching the membrane via micro-pipette aspiration. In this method, giant unilamellar vesicles, tens of micrometers in size, are created via electro-formation. Phosphatidyl Choline lipid is spread on an indium tin oxide substrate and hydrated with water. It is then subjected to an alternating electric field with its amplitude and frequency set to optimize the quantity and quality of vesicles. These vesicles are then studied under an optical microscope, and tension is applied using a micro-pipette with tip size in the micrometer scale. These micro-pipettes are pulled from 1 millimeter capillary tubes and coated with Silane to avoid vesicles sticking to them. They are attached to a water reservoir for the purpose of creating minute suction pressure by adjusting the water height in the reservoir. Aspirating a part of a vesicle disturbs its spherical symmetry and subjects the membrane to tension. A suitable dye which preferentially partitions into one of the co-existing phases is used in the vesicle sample. Under tension, dark patches are observed on the vesicles, implying that the regions are in different phases. A quantitative analysis of the phase separation process needs to be carried out to understand the microscopic origin of tension-induced phase separation.

2. Objective

To form giant unilamellar vesicles and study tension induced domain formation when subjected to micro-pipette aspiration.

3. Materials and Methods

3.1. Vesicle preparation

3.1.1 Lipid solution preparation

1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC) is the lipid of choice for this particular set of experiments. The second component in this binary system of lipids is 27-hydroxycholesterol. Lipid solution is made in Chloroform and is mixed with DiI, a fluorescent lipophilic cationic indocarbocyanine dye, at a molar percentage of 0.5. The choice of the dye is due to its property to fluoresce weakly in water but fluoresce highly and become quite photostable when incorporated into membranes. The initial choice of dye Rhodamine DHPE had to be discarded because of its photoreaction which results in similar observation of domains even under tensionless scenario.

3.1.2 Electroformation

The vesicles are formed via process called electroformation. A chamber is prepared over a glass plate which is coated with Indium Tin Oxide (ITO) to make the plate electrically conducting, and a cylinder of Teflon is used for the containment of the sample while electroformation. A thin coating of the lipid sample is made on the bottom surface of the chamber, and the chamber is filled with a suitable aqueous solution. In the initial stages of investigation, sucrose solution was chosen for the ease of observation as when the vesicles are formed with sucrose solution inside and studied in dextrose solution, the vesicles will sink to bottom of observation chamber. These solutions are prepared in such a way that the concentration of the sucrose solution inside the vesicles is slightly higher than that of the dextrose solution environment in the observation chamber. This is intended to reduce the tension in vesicle membrane via osmosis in outward direction.

In the later stages of investigation, vesicles were prepared and observed in pure water to confirm the validity of the observations under normal conditions. The tension reduction was achieved by preparing the vesicles in slightly higher temperature than during observation. The internal volume reduction caused by reduction in temperature of vesicle system during observation facilitates reduction in membrane tension.

The coated lipid sample together with the desired solution is subjected to an alternating electric field for a duration of two hours. This alternating electric field causes disturbance in the polar aqueous medium and exposes the lipids to water, forcing them to form bilayers and thereby vesicles. Lipids, having two tails, are inclined to form bilayers to reduce tail interaction with water, unlike single tailed soap molecules forming micelles or molecules with higher number of tails which adopt inverse hexagonal packing. It has been found that the quantity and quality of vesicles are optimal when the alternating electric field is set to frequency of 12 Hertz and root mean square amplitude of 600 millivolts.

The sample is then subjected to a low amplitude direct electric field to make sure that the vesicles are not sticking to the bottom of the chamber.

3.2. Micro-pipette forging

The vesicles are found to be in micron scale, around tens of micrometers in size. To apply tension on these membranes via aspiration, micro-pipettes with tip size in similar order need to be forged.

3.2.1 Micropipette crafting

These micropipettes are pulled from capillary tubes of 1mm inner diameter after cleaning with ethanol and dried. The middle segment of the capillary tube is heated while the ends are pulled apart so as to create a sharp and narrow segment. The pulling closes the capillary tube at this segment and the closing needs to be cut off. A bead of glass with melting temperature lesser than the material of the capillary tube is made on a platinum electrode. When the bead is heated to the point when the bead becomes fluid, it expands in volume. The closed tip of the capillary is allowed to touch the fluid bead and the bead is let to cool. When the bead cools down, it contracts, breaking and taking away with it the portion of the capillary tube that was touching the bead. Now that the tip is open, the concern shifts to making the pipette with desired tip size. The pulled region, which is almost 1 centimeter in length, is a long conical shape, with the height to base radius ratio high enough that the walls can be considered to be parallel in micron scale. The glass bead is heated again to the fluid state, and the tip of the pipette is allowed to touch the bead. The capillary action makes the fluid glass creep into the capillary tube slowly. Once the meniscus reaches the point where the radius is in desired range, the bead is allowed to cool. The molten length of glass inside the pipette cools down, the height of the so formed cylindrical pillar undergoes linear contraction, cutting and pulling away the portion of the pipette which was filled by the molten glass.

3.2.2 Fire polishing

The above step is in effect breaking away a portion of the glass capillary, leaving sharp, possibly uneven, edges on the pipette tip, which can be harsh for vesicles and can burst them. To troubleshoot this problem, each pipette undergoes fire polishing. A clean platinum electrode is heated and the tip of pipette is brought near the electrode. The heat radiated gently melts the sharp edge of the tip and polishes the pipette tip. This step needs attention as the heat deforms the glass capillary tube and shrinks the tip in radius making the pipette lose its uniformity in radius, and can potentially close the tip.

3.2.3 Silanization

The final step in micropipette forging is to silanize the pipettes to avoid vesicles sticking to it. Silan ((3-Mercaptopropyl)-trimethoxysilan) is a silicon hydrocarbon which is hydrophilic in nature. The pipette tips are coated with 2 percentage silane solution in ethanol by aspiration. Excess silane is removed by aspirating ethanol and releasing it. The pipettes are then subjected to high temperature for almost two hours for the ethanol to evaporate and the silan coating to settle down on the walls. Silanization is done multiple times for a single pipette for better silan coating.

3.2.4 Setup

For the experimental setup, the pipette is attached to a water reservoir via a flexible connection pipe. The height of the water reservoir is changed to bring about a hydraulic change in pressure. This technique is found to be reliable with the least possible step of change in water height as 5 micrometers. The control of water reservoir height is managed by a computer controlled vertical transition stage. Observations were carried out using an inverted optical microscope.

3.3. Analysis

The micropipette is filled with the same solution in the observation chamber to maintain uniformity in the environment throughout the analysis as a different solution filled in the pipette can cause disruption in uniformity during outflow. Precaution should be taken to remove air bubbles from pipette before analysis as they can cause huge margin of error in pressure application. Tension is applied on the membrane by aspirating a part of membrane into the micropipette. A vesicle of desired size is chosen and the micropipette is brought near the vesicle. The height of water reservoir is gently decreased to create small increase in suction pressure at the pipette tip. A fine micromanipulator mounted to a course micromanipulator is used for smooth movement of micropipette.

4. Observations

In the initial stages of the investigation, the vesicles were prepared in 100 millimolar concentration of sucrose solution. These vesicles were then studied in an environment of 115 millimolar concentration of dextrose solution. The concentration difference between inside and outside the vesicles were found sufficient to reduce membrane tension before aspiration. Most of the vesicles were found at the bottom of the observation chamber. These vesicles were binary systems of POPC lipid and 27-hydroycholesterol, along with a dye Rhodamine DHPE. Before aspiration, both these components are found to be existing in liquid disordered phase. Under application of tension via micropipette aspiration, it is observed that the vesicle membrane undergoes a phase transition, leading to the coexistence of the two phase. The dye used in the lipid sample was chosen because of its property to dissolve only in liquid disordered phase. This property gives rise to dark patches wherever the oxysterol rich phase is present.

Fig 1 Micropipette aspiration of a vesicle (Fluorescence imaging and Differential Interference Contrast imaging)

These small numerous subdomains then move around the vesicle and come together to coalesce to find larger domains in order to decrease the line energy. All the smaller domains in a vesicle come together to form a single huge domain after a sufficient duration of time. The change in geometric shape that follows depends on the volume ratio of the oysterol in the vesicle membrane. Homogeneity in composition of lipid and oxysterol in each vesicle is not guaranteed by homogeneity of lipid sample initially prepared. During vesicle formation, some vesicles will have more number of oysterol molecules in the membrane than others, some vesicles may be created with little oxysterol.

Fig 2 Oxysterol domain formation in POPC lipid vesicle prepared in 100mM Sucrose solution, observed in 115mM Dextrose solution

If the final domain formed is smaller than half the size of the vesicle, the vesicle is lipid dominated. The composition of oxysterol and lipids is in the desired range. Under micropipette aspiration, phase separated domains are observed in short period of time after aspiration. These domains keep traversing around the vesicle and the aspirated tether. They reduce in number and increase in area by coalescing with other domains until they become one single huge domain. This domain is suspected to be in liquid ordered phase. Under Differential Interference Contrast imaging and Phase Contrast imaging, it is observed that these domains form bulges in the spherical geometry of vesicles when the tension is immediately removed, while the tension during aspiration prevents the domains from forming bulges due to membrane stretching. These bulges are caused by the line tension between the coexisting phases.

Fig 3 Domain formation observed during aspiration and after immediate release (Fluorescence imaging)

Fig 4 Phase Contrast imaging of aspirated vesicle showing no fluctuation after phase separation

Fig 5 Phase Contrast imaging of vesicle showing fluctuations after immediate release after phase separation

On the other hand if the final domain formed is larger than half of size of the vesicle, the vesicle is said to be oxysterol dominated. Under tension, the dark domain formed will constitute more area in comparison to liquid disordered phase region. In this case, unlike forming bulges on surface of vesicle, it is observed that the new phase dominates the vesicle while the liquid disordered phase region seems to bulge out. Upon further increase in membrane tension, the dark domain undergoes more aspiration, reducing the surface area of the non-aspirated spherical region, and the liquid disordered domains are found to form individual vesicles which remain attached to the dark vesicle. Under high tension, these newly formed individual vesicles are found to be stationary unlike the domains moving around the vesicle in low tension. Under reduction of tension, these individual vesicles coalesce with the aspirated vesicle to form a single vesicle with liquid disordered phase domains. This phenomenon of formation of independent liquid disordered phase vesicles under high tension is found to be reversible under reduction of tension, and reproducible under further increase in tension.

Fig 6 Sterol dominated vesicle undergoing phase separation, causing liquid disordered phase region to form individual vesicle

Fig 7 Phase Contrast imaging confirming formation of individual vesicles and closure of bilayer

A third case is when the vesicle was formed with too little oxysterol in the bilayer. The number of oxysterol molecules will be too less to segregate together and to form domains of noticeable size. In this case, domains are not observed even under high tension.

Fig 8 Aspiration of vesicle with insufficient amount of oxysterol resulting in observation of no domain formation

In the later stages of investigation, the sucrose and dextrose solutions were substituted by filtered water for studying the validity of the phenomenon under normal conditions and to decouple any influence of the solutions of previous choice. The vesicles were prepared in filtered water under both sine and square wave alternating electric fields. These vesicles were then analyzed in filtered water. The observation of domain formation confirmed the validity of the phenomenon.

Fig 9 Domain formation under applied tension in vesicle prepared in filtered water (Sample 1)

Fig 10 Domain formation under applied tension in vesicles prepared in filtered water (Sample 2)

Fig 11 Domain formation under applied tension in vesicles prepared in filtered water (Sample 3)

The use of the dye Rhodamine DHPE posed problems in the experiment since this dye exhibits photobleaching under illumination. Vesicles with Rhodamine DHPE were observed under 60x magnification under microscope. Two interesting phenomena were observed.

1. Domain formations in non-tensed vesicles. A set of study was carried out without tensing the vesicle membrane. Without application of tension, the vesicles were studied under 60x magnification in an optical microscope. The fluctuations were observed in membrane through Differential Interference Contrast imaging to guarantee that the membranes experienced no tension. In a short period of time, the vesicles started exhibiting domains. Even after the domains were formed, the membrane was not found to be completely tense. This was observed to be irreversible in nature. After leaving the vesicles in dark for considerable amount of time, the vesicles didn’t retrieve homogeneity.

Fig 12 Domain formation without application of tension

Fig 13 Phase Contrast imaging showing contours in vesicle to show membrane relaxation

2. Reduction in intensity of fluorescence. Some vesicles don’t exhibit domains even under tension. When there is insufficient amount of oxysterol on the membrane, it is difficult for them to segregate and form domains of noticeable dimensions. These vesicles don’t show domains even under application of high membrane tension. When these vesicles are subjected to illumination under 60x magnification in an optical microscope without application of membrane pressure, the fluorescence intensity of the vesicles are observed to decrease with passing of time. The gradual decrease in the fluorescence intensity leads to complete loss of fluorescence of the dye. The dye gets relieved off its fluorescence property and thus the vesicle turns dark eventually.

Fig 14 Unaspirated vesicle under illumination 160 seconds apart

Fig 15 Intensity plot of above shown vesicle under illumination 160 seconds apart

Due to these difficulties, the dye of choice was changed from Rhodamine DHPE to DiI, a fluorescent lipophilic cationic indocarbocyanine dye which has a higher tolerance against photobleaching. Rhodamine DHPE was replaced with DiI in the lipid sample solution with a molar percentage of 0.5. The new choice of dye seems to have successfully reduced margin of error caused by previous choice of dye.

Fig 16 DiI dyed vesicles under aspiration (Sample 1)

Fig 17 DiI dyed vesicles under aspiration (Sample 2)

To tackle the problem of non-uniform distribution of oxysterol among vesicles, various approaches were adopted. The inhomogeneity in composition of the lipid-oxysterol solution is something that is not controllable during electroformation. The oscillating electric field disturbs the polar medium around the lipids and forces them to form bilayers and thereby vesicles. During these disturbances, both lipid and oxysterol molecules move across to attach to nearest bilayer. In this random motion, homogeneity in bilayer composition is not guaranteed.

A new method of study was adopted in which oxysterols are delivered to vesicles from the solution medium. The lipid sample was prepared without oxysterol and vesicles were prepared from this lipid sample in oxysterol-water solution. Solubility of 27-hydroxycholesterol in water is found to be minimal and close to insoluble. Solutions of oxysterol in water at concentration of few tens of micrograms per milliliter of water were used for preparing vesicles as well as for solution environment during analysis.

The objective was to study absorption of oxysterol molecules from the aqueous medium to the membrane. The vesicles were almost devoid of oxysterol molecules as the lipid solution was prepared without oxysterol. The only oxysterol molecules that could incorporate into the membrane came from the aqueous solution of oxysterol used for vesicle preparation. More oxysterol molecules got eventually incorporated into the membrane from the oxysterol solution used for environment during the stage of analysis.

The upper limit of oxysterol solution concentration is governed by ease of experimenting. At concentration of 0.05 mg/mL of oxysterol, the vesicles were found sticking to each other and forming clusters. The linkage property of sterol causes vesicles to share membrane and stick to each other, forming giant clusters. These clusters pose problem in isolating and aspiration single vesicles. Excess sterol concentration also increases fluidity of lipids, providing them capability of long distance stretching. This causes problem while aspirating isolated vesicles as they form channels which provide lipids to the aspirated vesicle and reduces tension experienced by the membrane.

Oxysterol solution concentration of 0.025 mg/mL is found to be delivering sufficient amount of oxysterol molecules to the vesicle membranes while not disturbing the identity of isolated vesicles. Domain formations were observed and analyzed and found to be in accordance with expected behaviour.

Fig 18 Domain formation observed in POPC vesicle prepared and analyzed in 0.025mg/mL 27-hydroxycholesterol

Fig 19 Domain formation and formation of individual vesicles of different phases under tension in 0.025mg/mL 27-hydroxycholesterol

Fig 20 Domain formation and formation of individual vesicles of different phases under tension in 0.025mg/mL 27-hydroxycholesterol

Fig 21 Domain formation and formation of individual vesicles of different phases under tension in 0.025mg/mL 27-hydroxycholesterol

5. Conclusion

The phenomenon of domain formation in lipid vesicles of binary component system under tension was studied. These domains are assumed to differ from the initial liquid disordered phase region only in the composition of lipid and oxysterol molecules, which alters solubility of dye in the region. In a lipid dominated vesicle under tension, the domains are refrained from forming bulges on the vesicle surface by the stretching of the membrane. In domain dominated vesicles under tension, despite the membrane stretching, the known liquid disordered phase forms bulges on the vesicle surface until a certain membrane tension above which these lipid patches form independent vesicles. More study on the phenomenon can show more light into the characteristics and nature of these domains.

References

1. J.R. Henriksen and J.H. Ipsen, ​Measurement of membrane elasticity by micro-pipette aspiration, Eur Phys J E Soft Matter, 14(2):149-67, 2004

​2. J. Mazej, Phase Transitions in Lipid Membranes​, Physics LibreTexts, 1-3, 2009

​3. V. Heinrich and W. Rawicz, Automated, high-resolution micropipet aspiration reveals new insight into the physical properties of fluid membranes, 21(5):1962-71, 2005

4. ​M. L. Longo and H. V. Ly, Micropipet aspiration for measuring elastic properties of lipid bilayers, Methods Mol Biol, 400:421-37, 2007